THE MUTATIONS OF OENOTHERA STENOMERES IO9 
dwarfs, and one corresponded to mut. lasiopetala from f. typica. It 
may be called Oe. stenomeres mut. gigasi lanosai. Not only were the 
petals hairy, as in mut. lasiopetala, but the filaments of the stamens 
alternate with the petals were densely la nose. The filaments of the 
other four stamens were glabrous. This plant was just coming into 
flower at the time of the first killing frost. It had to be removed to 
the greenhouse and does not give much promise of yielding seeds. 
The plants which failed to flower appeared to have the gigas- 
characteristics. It will be remembered that Oe. gigas de Vries also 
bred true from the first appearance of the mutation. 
Conclusion 
The point which should be especially emphasized in connection 
with the mutations of Oe. stenomeres is that neither of them can be 
interpieted as the result of segregation following hybridization. Oe. 
stenomeres is an isolated cruciate species, related to none of the other 
cruciate species, and geographically remote from all of them. The 
peculiarities which characterize mut. lasiopetala are not known in any 
othei species and cannot, therefore, be construed as having been 
introduced into Oe. stenomeres by hybridization with any other type. 
It seems more likely that Oe. stenomeres itself originated by mutation 
from some broad-petaled elementary species allied to Oe. gauroides 
Hornem., and that mut. lasiopetala constitutes a second mutation in 
the same direction. The occurrence of mut. gigas cannot be viewed 
as even remotely due to hybridization, for the characters of this 
mutation can be explained on a cytological basis. 
Like most of the other Onagrae, Oe. stenomeres fhows a considerable 
degree of pollen sterility. This sterility we may view as a sign of the 
prevalence of mutation in the group. Both mutation and hybridiza- 
tion bring about pollen sterility, and both processes are widely pre- 
valent in the subgenus Onagra. It is illogical and unjustifiable to 
state, as certain authors have done, that the mutability is the result 
of the hybridity. It seems far more probable that the truth is exactly 
the opposite. Far from having proved that hybridization "explains" 
mutation, no one has yet even shown how hybridization cotdd explain 
it. Finally, attention may again be called to the fact that Oe. steno- 
meres is a close-pollinated and practically a cleistogamous species in 
nature, and that its mutability was not discovered until the fourth 
generation of a guarded line was grown. 
Bureau of Plant Industry, Washington 
